Computer Vision Engineer

We’re looking for a Computer Vision Engineer to join a cutting-edge team driving R&D innovation in radar-based systems. In this role, you will lead the end-to-end development of Computer Vision models, from data strategy and model training to deployment in real-world defense applications.

Key Responsibilities

• Algorithm Development: Design and implement robust algorithms for real-time detection, classification, and tracking of objects on high-speed conveyor systems.
• End-to-End Pipeline: Develop CV pipelines that combine classic image processing techniques with state-of-the-art Deep Learning models.
• Edge Optimization: Port and optimize models for the NVIDIA Jetson platform, ensuring maximum throughput and minimal power consumption.
• System Integration: Collaborative development of interfaces between vision software and robotic control systems.
• Data Reliability: Build tools for automated data collection, labeling, and model validation to ensure performance remains stable in changing industrial environments.

Qualifications

• Experience: 4+ years of professional experience in Computer Vision and algorithms development.
• Computer Vision Fundamentals: Deep mastery of Classic CV (image filtering, feature extraction, geometry) alongside modern Deep Learning.
• Model Specialization: Proven track record in Object Detection, Image Classification, and Object Tracking (e.g., SORT/DeepSORT, ByteTrack) in dynamic environments.
• Software Toolkit: Expert-level proficiency in Python, PyTorch, and OpenCV.
• Edge & Hardware: Experience with Edge Deployment (specifically NVIDIA Jetson) and a strong understanding of Linux/Arm architectures.

Preferred Qualifications

• Programming: Proficiency in C / C++ for performance-critical modules.
• Environment: Strong command of Linux, Docker, and Arm-based systems.
• Audio Signal Processing: Implement ML pipelines for audio feature extraction and classification.
• Streaming: Knowledge of Video Streaming protocols and low-latency video processing (RTSP, GStreamer).